| 1 |
#!/usr/bin/perl |
| 2 |
use strict; |
| 3 |
|
| 4 |
BEGIN { require 'common.pl' } |
| 5 |
|
| 6 |
require Encode; |
| 7 |
|
| 8 |
my $max_result = 30; |
| 9 |
|
| 10 |
sub decode_url ($) { |
| 11 |
my $s = shift; |
| 12 |
$s =~ tr/+/ /; |
| 13 |
$s =~ s/%([0-9A-Fa-f]{2})/pack 'C', hex $1/ge; |
| 14 |
return Encode::decode ('utf-8', $s); |
| 15 |
} # decode_url |
| 16 |
|
| 17 |
sub htescape ($) { |
| 18 |
my $s = shift; |
| 19 |
$s =~ s/&/&/g; |
| 20 |
$s =~ s/</</g; |
| 21 |
$s =~ s/>/>/g; |
| 22 |
$s =~ s/"/"/g; |
| 23 |
return $s; |
| 24 |
} # htescape |
| 25 |
|
| 26 |
my $param = {}; |
| 27 |
for (split /[&;]/, $ENV{QUERY_STRING} || '') { |
| 28 |
my ($name, $value) = split /=/, $_, 2; |
| 29 |
$param->{decode_url ($name)} = decode_url ($value); |
| 30 |
} |
| 31 |
|
| 32 |
my $pattern = quotemeta normalize $param->{word}; |
| 33 |
my $eword = htescape $param->{word}; |
| 34 |
|
| 35 |
print qq[Content-Type: text/html ; charset=utf-8 |
| 36 |
|
| 37 |
<!DOCTYPE HTML> |
| 38 |
<html lang=en> |
| 39 |
<head> |
| 40 |
<title>Search result for "$eword"</title> |
| 41 |
<link rel=stylesheet href="/www/style/html/xhtml"> |
| 42 |
<style> |
| 43 |
mark { |
| 44 |
background-color: yellow; |
| 45 |
} |
| 46 |
.tag { |
| 47 |
color: maroon; |
| 48 |
} |
| 49 |
.ref { |
| 50 |
color: orange; |
| 51 |
} |
| 52 |
.rfc2119 { |
| 53 |
font-weight: bolder; |
| 54 |
} |
| 55 |
.pattern-star { |
| 56 |
font-style: normal; |
| 57 |
color: gray; |
| 58 |
text-decoration: none; |
| 59 |
border-style: none; |
| 60 |
} |
| 61 |
.pattern-var { |
| 62 |
font-style: italic; |
| 63 |
color: gray; |
| 64 |
text-decoration: none; |
| 65 |
border-style: none; |
| 66 |
} |
| 67 |
</style> |
| 68 |
</head> |
| 69 |
<body> |
| 70 |
<h1>Search result for "$eword"</h1>]; |
| 71 |
|
| 72 |
my $has_match; |
| 73 |
for_each_data_file (sub ($) { |
| 74 |
my $data_file_name = shift; |
| 75 |
load_data_file ($data_file_name, my $exact_data = {}, my $pattern_data = {}); |
| 76 |
$pattern_data = unescape_patterns ($pattern_data); |
| 77 |
$has_match |= print_matches ($data_file_name, $exact_data, $pattern_data); |
| 78 |
}); |
| 79 |
|
| 80 |
unless ($has_match) { |
| 81 |
print q[<p>No match found.]; |
| 82 |
} |
| 83 |
|
| 84 |
sub unescape_patterns ($) { |
| 85 |
my $pattern_data = shift; |
| 86 |
my $new_data = {}; |
| 87 |
for (keys %$pattern_data) { |
| 88 |
my $w = $_; |
| 89 |
my $v = $_; |
| 90 |
$v =~ s/\(\.\+\)/*/g; |
| 91 |
$v =~ s/\\([\s\S])/$1/g; |
| 92 |
$new_data->{$v} = $pattern_data->{$w}; |
| 93 |
} |
| 94 |
return $new_data; |
| 95 |
} # unescape_patterns |
| 96 |
|
| 97 |
sub print_matches ($$$) { |
| 98 |
my ($file_name, $exact_data, $pattern_data) = @_; |
| 99 |
|
| 100 |
my $file_id = $file_name; |
| 101 |
if ($file_name =~ /([0-9A-Za-z-]+)\.dat$/) { |
| 102 |
$file_id = $1; |
| 103 |
} |
| 104 |
|
| 105 |
my $has_match; |
| 106 |
my $r = qq[<h2>File "<a href="edit/@{[htescape $file_id]}"><code class=file>@{[htescape $file_id]}</code></a>"</h2><table>]; |
| 107 |
my $result = 0; |
| 108 |
my $added = {}; |
| 109 |
my %en2ja = (%$exact_data, %$pattern_data); |
| 110 |
for (sort {$a cmp $b} keys %en2ja) { |
| 111 |
if (/$pattern/) { |
| 112 |
$has_match = 1; |
| 113 |
$added->{$_} = 1; |
| 114 |
$r .= get_match_text ($_, $en2ja{$_}, $pattern); |
| 115 |
$r .= q[<tr><td colspan=2>...] and last if ++$result == $max_result; |
| 116 |
} |
| 117 |
} |
| 118 |
$result = 0; |
| 119 |
my %ja2en = reverse %en2ja; |
| 120 |
for (sort {$a cmp $b} keys %ja2en) { |
| 121 |
if (/$pattern/) { |
| 122 |
next if $added->{$ja2en{$_}}; |
| 123 |
$has_match = 1; |
| 124 |
$r .= get_match_text ($ja2en{$_}, $_, $pattern); |
| 125 |
$r .= q[<tr><td colspan=2>...] and last if ++$result == $max_result; |
| 126 |
} |
| 127 |
} |
| 128 |
$r .= q[</table>]; |
| 129 |
print $r if $has_match; |
| 130 |
return $has_match; |
| 131 |
} # print_matches |
| 132 |
|
| 133 |
sub get_match_text ($$) { |
| 134 |
my ($en, $ja) = @_; |
| 135 |
|
| 136 |
## NOTE: Marking will not work well if it contains &, <, >, or ", or |
| 137 |
## the pattern matches with charrefs, e.g. "t" (part of < and "). |
| 138 |
|
| 139 |
my $r = q[<tr><td lang=en>]; |
| 140 |
my $v = htescape ($en); |
| 141 |
$v =~ s[($pattern)][<mark>$1</mark>]g; |
| 142 |
$v =~ s[(<[\s\S]+?>)][<span class=tag>$1</span>]g; |
| 143 |
$v =~ s[(&[#0-9A-Za-z]+;)][<span class=ref>$1</span>]g; |
| 144 |
$v =~ s[\*][<var class=pattern-star>*</var>]g; |
| 145 |
$r .= $v; |
| 146 |
$r .= q[<td lang=ja>]; |
| 147 |
my $v = htescape ($ja); |
| 148 |
$v =~ s[($pattern)][<mark>$1</mark>]g; |
| 149 |
$v =~ s[(<[\s\S]+?>)][<span class=tag>$1</span>]g; |
| 150 |
$v =~ s[(&[#0-9A-Za-z]+;)][<span class=ref>$1</span>]g; |
| 151 |
$v =~ s{(\[\[[^\[\]]+\]\])}[<span class=rfc2119>$1</span>]g; |
| 152 |
$v =~ s[(\$[0-9]+)][<var class=pattern-var>$1</var>]g; |
| 153 |
$r .= $v; |
| 154 |
|
| 155 |
return $r; |
| 156 |
} # get_match_text |
Parent Directory
| 
Revision Log